
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from datetime import datetime
from fbprophet import Prophet
from IPython.display import display,HTML
%matplotlib inline

Importing plotly failed. Interactive plots will not work.


df = pd.read_csv("AC_Data.csv").dropna()
df['Date'] = pd.to_datetime(df['0'])
df = df.drop('0',axis =1).set_index('Date')
display(HTML(df.head().to_html()))


sns.pairplot(df)

<seaborn.axisgrid.PairGrid at 0x7feaa536ce80>


import requests

url = "https://visual-crossing-weather.p.rapidapi.com/forecast"

querystring = {"aggregateHours":"24","location":"Washington,DC,USA","contentType":"csv","unitGroup":"us","shortColumnNames":"0"}

headers = {
    'x-rapidapi-host': "visual-crossing-weather.p.rapidapi.com",
    'x-rapidapi-key': "04105d6946mshac13b0cd8141cf4p128e31jsne96aca6920d7"
    }

response = requests.request("GET", url, headers=headers, params=querystring)

print(response.text)

{"message":"You are not subscribed to this API."}


sum_df = pd.DataFrame(df.sum(),columns = ['name']).sort_values(by='name')


corr = df.corr()
corr.style.background_gradient().set_precision(4)


print('-'*30)

print("Least Used AC: ",sum_df.index[0])
print("Total amount: ",int(sum_df.loc[sum_df.index[0]][0]))

print('-'*30)

print("Most Used AC: ",sum_df.index[-1],"")
print("Total amount: ",int(sum_df.loc[sum_df.index[-1]][0]))

print('-'*30)

------------------------------
Least Used AC:  AC 5
Total amount:  50800
------------------------------
Most Used AC:  AC 18 
Total amount:  223573
------------------------------


# from datetime import datetime
# import matplotlib.pyplot as plt
# #from meteostat import Point, Daily

# # Set time period
# start = datetime(2018, 1, 1)
# end = datetime(2018, 12, 31)

# # Create Point for Vancouver, BC
# #location = Point(49.2497, -123.1193, 70)

# # Get daily data for 2018
# data = Daily(49.2497, -123.1193, start, end)
# data = data.fetch()


fig, ax = plt.subplots(figsize = (15,10))
plt.bar(sum_df.index, sum_df.name, color='#a8dadc', width=0.8)
for i, v in enumerate(sum_df.name):
    ax.text(i-len(sum_df.name)*0.024, v+1500, str(int(v)))
plt.title('Total AC Usage across time-period')
plt.xlabel('AC Name')
plt.ylabel('Total amount of usage')
plt.show()


daily_sum = df.groupby(pd.Grouper(freq="D")).sum()
weekly_sum = df.groupby(pd.Grouper(freq="W")).sum()


x_week = [i.isocalendar()[1] for i in weekly_sum.index]


x_day = [i.date().strftime('%Y-%m-%d') for i in daily_sum.index]


fig, ax = plt.subplots(figsize = (15,10))
for ac in list(weekly_sum):
    plt.plot(x_week,weekly_sum[ac] ,label = ac)
plt.legend(loc="upper left")
plt.title('Weekly AC usage trends')
plt.xlabel('Week')
plt.ylabel('Total amount of usage')
plt.xticks(x_week,rotation = 70)
plt.show()


fig, ax = plt.subplots(figsize = (15,10))
plt.plot(x_day,daily_sum['AC 18'] ,label = 'AC 18')
plt.plot(x_day,daily_sum['AC 16'] ,label = 'AC 16')
plt.legend(loc="upper left")
plt.title('Correlation between AC 16 and AC 18')
plt.xlabel('Week')
plt.ylabel('Total amount of usage')
plt.xticks(x_day,rotation=70)
plt.show()


fig, ax = plt.subplots(figsize = (15,10))
plt.plot(x_day,daily_sum['AC 1'] ,label = 'AC 1')
plt.plot(x_day,daily_sum['AC 2'] ,label = 'AC 2')
plt.legend(loc="upper left")
plt.title('Correlation between AC 1 and AC 2')
plt.xlabel('Week')
plt.ylabel('Total amount of usage')
plt.xticks(x_day,rotation=70)
plt.show()


def get_df_forecast_model(df_name,col_name):  
    df_to_forecast = pd.DataFrame()
    df_to_forecast['y'] = df_name[col_name]
    df_to_forecast['ds'] = df_name.index
    df_to_forecast.reset_index(inplace = True)
    df_to_forecast.drop('Date',inplace = True ,axis = 1)
    forecast_model = Prophet(interval_width=0.95)
    forecast_model.fit(df_to_forecast)
    return(forecast_model)


def show_forecast_results(forecast_model,name):
    print('*'*100)
    print('-'*100)
    _forecast = forecast_model.make_future_dataframe(periods=10, freq='D')
    _forecast = forecast_model.predict(_forecast)
    plt.figure(figsize=(18, 6))
    forecast_model.plot(_forecast, xlabel = 'Date', ylabel = name)
    plt.title('10 days Forecast for '+name)

    fig, ax = plt.subplots(figsize = (15,10))
    known, = plt.plot(_forecast['ds'].iloc[:-10] ,_forecast['yhat'].iloc[:-10],label = name+'  Known Usage')
    forecasted, = plt.plot(_forecast['ds'].iloc[-10:] ,_forecast['yhat'].iloc[-10:],label = name + ' forecast usage')
    plt.legend(loc="upper left")
    plt.title('10 Days Forecast for '+ name)
    plt.xlabel('Date')
    plt.ylabel('Total amount of Daily usage')
    plt.xticks(_forecast['ds'],rotation=70)
    plt.show()
    print('-'*15," Forcasted Values for ",name," ", '-'*15)
    forecasting_df_only_values = (_forecast[['ds','yhat']].iloc[-10:])
    forecasting_df_only_values['Date'] = forecasting_df_only_values['ds'].dt.date
    forecasting_df_only_values['Usage'] = forecasting_df_only_values['yhat']
    forecasting_df_only_values.drop(['ds','yhat'],inplace = True,axis = 1)
    style_df = forecasting_df_only_values.style.background_gradient()
    display(HTML(style_df.render()))
    #jh_forecast.iloc[:-11].style.background_gradient()


for ac in df:
    f_model = get_df_forecast_model(daily_sum,ac)
    show_forecast_results(f_model,ac)

INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.
INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.

****************************************************************************************************
----------------------------------------------------------------------------------------------------

<Figure size 1296x432 with 0 Axes>

---------------  Forcasted Values for  AC 1   ---------------

INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.
INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.

****************************************************************************************************
----------------------------------------------------------------------------------------------------

<Figure size 1296x432 with 0 Axes>

---------------  Forcasted Values for  AC 2   ---------------

INFO:fbprophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.
INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.

****************************************************************************************************
----------------------------------------------------------------------------------------------------

	AC 1	AC 2	AC 3	AC 4	AC 5	AC 6	AC 7	AC 8	AC 9	AC 10	AC 11	AC 12	AC 13	AC 14	AC 15	AC 16	AC 17	AC 18
Date
2019-08-01 00:00:00	7.518632	8.788315	0.0	0.0	2.617045	4.079041	2.782276	4.624447	5.222060	2.151238	1.585072	0.560373	3.142941	2.749470	5.417774	4.113460	3.305072	6.735981
2019-08-01 00:02:00	7.426114	8.940615	0.0	0.0	2.581625	3.781231	2.529366	5.057423	5.349465	2.414715	2.168184	1.818730	3.085110	2.720484	3.302422	3.986483	3.220588	6.379500
2019-08-01 00:04:00	7.052986	9.161103	0.0	0.0	2.592095	3.800127	2.332304	6.322521	3.995392	2.237114	3.345624	2.310409	3.132799	2.676861	3.539026	3.797881	3.131560	6.363475
2019-08-01 00:06:00	6.665446	9.065626	0.0	0.0	2.575639	3.772891	2.596200	5.805132	3.553778	1.878356	2.737645	2.510972	3.090007	2.666604	4.607439	3.062610	3.063953	6.127366
2019-08-01 00:08:00	6.674838	9.096130	0.0	0.0	2.021472	3.155697	2.845417	6.315060	3.042244	1.914650	2.452500	2.171791	3.097202	2.710572	5.332696	2.640150	2.864680	5.662474

	AC 1	AC 2	AC 3	AC 4	AC 5	AC 6	AC 7	AC 8	AC 9	AC 10	AC 11	AC 12	AC 13	AC 14	AC 15	AC 16	AC 17	AC 18
AC 1	1.0000	0.4160	0.2334	0.2811	0.0837	0.2040	0.0305	0.0997	0.1085	0.2002	0.3113	0.1846	0.2654	0.1808	0.3197	0.2812	-0.0931	-0.0357
AC 2	0.4160	1.0000	0.2131	0.3621	0.0940	0.1696	0.0679	0.0948	0.1471	0.1815	0.1943	0.1235	0.2379	0.1127	0.2917	0.2447	0.0395	-0.0432
AC 3	0.2334	0.2131	1.0000	0.2504	0.0565	0.0853	0.0047	0.0993	0.1250	0.0714	0.1433	0.1061	0.2131	0.1001	0.2701	0.2615	0.0256	0.0270
AC 4	0.2811	0.3621	0.2504	1.0000	0.1146	0.1487	0.0856	0.1605	0.1693	0.1850	0.1286	0.1503	0.2279	0.1826	0.1984	0.1779	0.0241	-0.0552
AC 5	0.0837	0.0940	0.0565	0.1146	1.0000	0.2255	0.1130	0.1180	0.1188	0.1211	0.1328	0.1053	0.0882	0.2019	0.0433	0.0479	0.0295	-0.0159
AC 6	0.2040	0.1696	0.0853	0.1487	0.2255	1.0000	0.1951	0.2107	0.1852	0.2062	0.2681	0.1803	0.2199	0.3051	0.1746	0.1626	-0.0502	-0.0914
AC 7	0.0305	0.0679	0.0047	0.0856	0.1130	0.1951	1.0000	0.1884	0.2126	0.2030	0.1694	0.2152	0.1299	0.2366	0.1214	0.0633	-0.0154	-0.0618
AC 8	0.0997	0.0948	0.0993	0.1605	0.1180	0.2107	0.1884	1.0000	0.1947	0.2050	0.1909	0.2344	0.1557	0.3364	0.1211	0.1314	-0.0963	-0.1246
AC 9	0.1085	0.1471	0.1250	0.1693	0.1188	0.1852	0.2126	0.1947	1.0000	0.2078	0.1757	0.1827	0.1871	0.2789	0.1725	0.1246	-0.0285	-0.0880
AC 10	0.2002	0.1815	0.0714	0.1850	0.1211	0.2062	0.2030	0.2050	0.2078	1.0000	0.1888	0.2042	0.1845	0.2929	0.2299	0.2495	-0.1796	-0.1435
AC 11	0.3113	0.1943	0.1433	0.1286	0.1328	0.2681	0.1694	0.1909	0.1757	0.1888	1.0000	0.2727	0.2071	0.3352	0.2560	0.2191	-0.1350	-0.1507
AC 12	0.1846	0.1235	0.1061	0.1503	0.1053	0.1803	0.2152	0.2344	0.1827	0.2042	0.2727	1.0000	0.1572	0.2540	0.1804	0.1574	-0.1279	-0.1079
AC 13	0.2654	0.2379	0.2131	0.2279	0.0882	0.2199	0.1299	0.1557	0.1871	0.1845	0.2071	0.1572	1.0000	0.2646	0.2196	0.2445	-0.0399	-0.0311
AC 14	0.1808	0.1127	0.1001	0.1826	0.2019	0.3051	0.2366	0.3364	0.2789	0.2929	0.3352	0.2540	0.2646	1.0000	0.1731	0.1807	-0.2038	-0.2183
AC 15	0.3197	0.2917	0.2701	0.1984	0.0433	0.1746	0.1214	0.1211	0.1725	0.2299	0.2560	0.1804	0.2196	0.1731	1.0000	0.3843	-0.1253	-0.1074
AC 16	0.2812	0.2447	0.2615	0.1779	0.0479	0.1626	0.0633	0.1314	0.1246	0.2495	0.2191	0.1574	0.2445	0.1807	0.3843	1.0000	-0.1129	-0.0749
AC 17	-0.0931	0.0395	0.0256	0.0241	0.0295	-0.0502	-0.0154	-0.0963	-0.0285	-0.1796	-0.1350	-0.1279	-0.0399	-0.2038	-0.1253	-0.1129	1.0000	0.6186
AC 18	-0.0357	-0.0432	0.0270	-0.0552	-0.0159	-0.0914	-0.0618	-0.1246	-0.0880	-0.1435	-0.1507	-0.1079	-0.0311	-0.2183	-0.1074	-0.0749	0.6186	1.0000

	Date	Usage
61	2019-10-01	1438.503798
62	2019-10-02	1771.153914
63	2019-10-03	1953.144364
64	2019-10-04	1880.209804
65	2019-10-05	1350.335550
66	2019-10-06	1215.799958
67	2019-10-07	1295.647470
68	2019-10-08	1478.901367
69	2019-10-09	1811.551482
70	2019-10-10	1993.541933

	Date	Usage
61	2019-10-01	2665.161648
62	2019-10-02	2383.077585
63	2019-10-03	2942.734990
64	2019-10-04	2612.774594
65	2019-10-05	2230.663584
66	2019-10-06	2383.981212
67	2019-10-07	2528.451559
68	2019-10-08	2734.778830
69	2019-10-09	2452.694767
70	2019-10-10	3012.352172

	Date	Usage
61	2019-10-01	1993.692191
62	2019-10-02	1832.350341
63	2019-10-03	1871.412607
64	2019-10-04	1999.633526
65	2019-10-05	1911.112543
66	2019-10-06	2033.390117
67	2019-10-07	1996.981535
68	2019-10-08	2101.921422
69	2019-10-09	1940.579572
70	2019-10-10	1979.641837

Importing Data

Initial Analysis using pairplot

Nothing can be concluded from these pair-plots¶

Tried a bunch of APIs to collect temperature data, but unsuccessful. Some did not work. Others were Paid. ¶

Correlation Matrix

We can observe that the usage of AC 17 is highly correlated with the usage of AC 18 and the usage of AC 1 is highly correlated with the usage of AC 2.¶

Most/Least Used AC

Weekly Usage Trends

The sudden dip in usage at week 33 could be due to the fact that hotel is closed for booking.¶

It can be noted that Usage is not absolute 0 for all ACs.¶

Correlation Trends for weekly usage

Forecasting Daily Usage Metrics for next 10 days

	Date	Usage
61	2019-10-01	1883.628845
62	2019-10-02	2029.767797
63	2019-10-03	2082.028877
64	2019-10-04	1920.277945
65	2019-10-05	1558.244717
66	2019-10-06	1561.337943
67	2019-10-07	1543.391180
68	2019-10-08	1981.813191
69	2019-10-09	2127.952143
70	2019-10-10	2180.213223

	Date	Usage
61	2019-10-01	1043.683531
62	2019-10-02	966.989137
63	2019-10-03	907.893297
64	2019-10-04	975.012622
65	2019-10-05	1099.555680
66	2019-10-06	1125.135112
67	2019-10-07	1071.585769
68	2019-10-08	1083.942603
69	2019-10-09	1007.248209
70	2019-10-10	948.152368

	Date	Usage
61	2019-10-01	1861.166088
62	2019-10-02	1792.022082
63	2019-10-03	1726.895495
64	2019-10-04	1809.530714
65	2019-10-05	1853.154834
66	2019-10-06	1965.424669
67	2019-10-07	1766.038743
68	2019-10-08	1920.006262
69	2019-10-09	1850.862256
70	2019-10-10	1785.735669

	Date	Usage
61	2019-10-01	2027.280761
62	2019-10-02	1883.856180
63	2019-10-03	1685.794305
64	2019-10-04	1669.476754
65	2019-10-05	1894.045996
66	2019-10-06	2080.750881
67	2019-10-07	2063.609575
68	2019-10-08	2074.012442
69	2019-10-09	1930.587861
70	2019-10-10	1732.525986

	Date	Usage
61	2019-10-01	2115.072496
62	2019-10-02	1813.714375
63	2019-10-03	1966.667741
64	2019-10-04	1872.550244
65	2019-10-05	1901.638652
66	2019-10-06	1822.614693
67	2019-10-07	2067.392823
68	2019-10-08	2214.110227
69	2019-10-09	1912.752105
70	2019-10-10	2065.705472

	Date	Usage
61	2019-10-01	2718.819609
62	2019-10-02	2376.486820
63	2019-10-03	2233.287892
64	2019-10-04	2067.357117
65	2019-10-05	2191.025870
66	2019-10-06	2640.238354
67	2019-10-07	2656.854441
68	2019-10-08	2845.900913
69	2019-10-09	2503.568123
70	2019-10-10	2360.369196

	Date	Usage
61	2019-10-01	1022.675369
62	2019-10-02	1030.498077
63	2019-10-03	935.778239
64	2019-10-04	964.088826
65	2019-10-05	956.057578
66	2019-10-06	1089.580999
67	2019-10-07	1012.889625
68	2019-10-08	1048.703302
69	2019-10-09	1056.526011
70	2019-10-10	961.806173

	Date	Usage
61	2019-10-01	1865.967663
62	2019-10-02	1977.204805
63	2019-10-03	1714.104071
64	2019-10-04	1778.511320
65	2019-10-05	1750.132994
66	2019-10-06	1841.470745
67	2019-10-07	1797.700528
68	2019-10-08	1948.679403
69	2019-10-09	2059.916545
70	2019-10-10	1796.815812

	Date	Usage
61	2019-10-01	1917.810296
62	2019-10-02	1732.408556
63	2019-10-03	1834.904688
64	2019-10-04	1829.691489
65	2019-10-05	1940.443907
66	2019-10-06	2038.792070
67	2019-10-07	1974.142922
68	2019-10-08	1979.150804
69	2019-10-09	1793.749064
70	2019-10-10	1896.245196

	Date	Usage
61	2019-10-01	2906.488396
62	2019-10-02	3033.390864
63	2019-10-03	3102.729056
64	2019-10-04	3043.324855
65	2019-10-05	2968.565065
66	2019-10-06	3221.806356
67	2019-10-07	2911.004744
68	2019-10-08	3025.295793
69	2019-10-09	3152.198261
70	2019-10-10	3221.536453

	Date	Usage
61	2019-10-01	2275.060888
62	2019-10-02	2252.293606
63	2019-10-03	2062.270102
64	2019-10-04	2058.767545
65	2019-10-05	2121.405042
66	2019-10-06	2274.715766
67	2019-10-07	2206.428030
68	2019-10-08	2377.046743
69	2019-10-09	2354.279461
70	2019-10-10	2164.255957

	Date	Usage
61	2019-10-01	2809.108043
62	2019-10-02	2693.737471
63	2019-10-03	2428.960398
64	2019-10-04	2471.340993
65	2019-10-05	2229.243032
66	2019-10-06	2826.923870
67	2019-10-07	2630.381203
68	2019-10-08	2857.735802
69	2019-10-09	2742.365230
70	2019-10-10	2477.588157

	Date	Usage
61	2019-10-01	1925.372722
62	2019-10-02	1843.134257
63	2019-10-03	1701.789644
64	2019-10-04	1767.894760
65	2019-10-05	1572.812516
66	2019-10-06	1710.472153
67	2019-10-07	1781.791943
68	2019-10-08	1962.993570
69	2019-10-09	1880.755104
70	2019-10-10	1739.410491

	Date	Usage
61	2019-10-01	3116.583916
62	2019-10-02	2969.192045
63	2019-10-03	2926.886014
64	2019-10-04	3020.020673
65	2019-10-05	3057.680025
66	2019-10-06	3306.462714
67	2019-10-07	3296.781266
68	2019-10-08	3256.634682
69	2019-10-09	3109.242810
70	2019-10-10	3066.936779

	Date	Usage
61	2019-10-01	4223.061050
62	2019-10-02	4041.532884
63	2019-10-03	4041.393613
64	2019-10-04	4274.424121
65	2019-10-05	4412.818637
66	2019-10-06	4966.141583
67	2019-10-07	4712.495383
68	2019-10-08	4372.136816
69	2019-10-09	4190.608650
70	2019-10-10	4190.469379